Two Concepts of Meditation and Three Kinds of Wisdom in Kamalaśīla’S Bhāvanākramas: a Problem of Translation

Buddhist Studies Review 23(1) 2006, 71–92 ISSN (print): 0256-2897 ISSN (online): 1747-9681

Two Concepts of Meditation and Three Kinds of Wisdom in Kamalaśīla’s Bhāvanākramas: A Problem of Translation

MARTIN T. ADAM Religious Studies Program, Department of Paciﬁ c and Asian Studies, University of Victoria, British Colombia, Canada [email protected]

ABSTRACT: A close reading of the three Bhāvanākramaḥ texts, written by Kamalaśīla (740–795 CE), reveals that their author was aware of two competing concepts of meditation prevalent in Tibet at the time of their composition. The two concepts of meditation, associated with the Sanskrit words bhāvanā and dhyāna, can be related respectively to the Indian and Chinese sides of the well-known debates at bSam yas. The account of the Mahāyāna path outlined in these texts implies an acceptance of the precedence of bhāvanā over dhyāna. In this paper I argue that Kamalaśīla advocated bhāvanā – a conception of meditation which encompasses non-conceptual dhyāna, but which also includes a discernment of reality (bhūta-pratyavekṣā) that is conceptual in nature. Such conceptual discernment should not be understood simply as a process of ordinary rational understanding (cintāmayī prajñā) but rather as constituting a special kind of meditative wisdom (bhāvanāmayī prajñā). A failure to recognize the subtle dif- ferences between Kamalaśīla’s employment of the terms dhyāna and bhāvanā, along with his advocacy of the latter, could easily lead to mistranslation and, with this, a basic misunderstanding of his position. In particular, it could lead to a conception of insight (vipaśyanā) that is overly intellectual in nature. Given the historically important role that these texts played in the formation of Tibetan Buddhism, the implications of such a misconception could be far-reaching. This paper attempts to clarify the key meditation terminology found in the Bhāvanākramas as well as demonstrate the rationale for using ‘meditation’ as the default translation for bhāvanā.

DIFFERENT CONCEPTS OF MEDITATION

The following pages contain an analysis of a constellation of meditation-related terms found in three polemical treatises, identically entitled Bhāvanākramaḥ (The Process of Meditation). It is not clear why their author, Kamalaśīla, wrote three texts with the same title.1 There is considerable overlap among the three, and not

1. The three texts were likely composed in Tibet between 792 and 794 CE, although see Taniguchi (1992) for an argument that the ﬁ rst Bhāvanākramaḥ was composed somewhat earlier in India.

infrequently repetition. That being said, the three treatises cover an extraordi- nary range of subjects, all united around the central purpose of providing guid- ance to new practitioners of the teachings of the Mahāyāna Sūtras.2 Historically, the Bhāvanākramas’ account of meditation has been enormously infl uential. Paul Williams has referred to the texts as ‘the principal systematic Indian sources for the integration of emptiness teachings into Madhyamaka meditation practice’ (Williams, 1989: 72). Elsewhere they have been described as ‘the origin of Tibetan tradition of how to meditate’ (Taniguchi, 1992: 303). This paper argues that there are, in fact, two competing concepts of meditation present in the texts. These two concepts are identifi able with two specifi c Sanskrit words, both of which have been commonly translated into English as ‘meditation’ – bhāvanā and dhyāna. Because Kamalaśīla does not employ these terms as syno- nyms, a problem arises for the modern day translator: which word, if either, should be privileged in translation as ‘meditation’? While neither can carry the same range of meanings as the English word (on which, see below), in this paper it is argued that Kamalaśīla regarded bhāvanā as normative for the practice of beginners in the way of the Mahāyāna Sutras. As such, ‘meditation’ should be its default translation. For these texts, it is potentially misleading to translate dhyāna as meditation. The issue is more than academic. Depending on the choice made, Kamalaśīla’s account of the Mahāyāna Buddhist path to Awakening will be radi- cally altered. To that extent, our understanding of both the doctrinal and practi- cal foundations of Tibetan Buddhism will be aff ected.

According to Edward Conze, ‘The fi rst explains the doctrine of the Mahāyāna, the second how it can be meditated upon, and the third what is the result of meditation’ (1975: 177). Conze is here following a description contained in a Tibetan record cited in Tucci (1958: 40–41). The account has it that the Tibetan king, Khri Srong lde btsan, requested these explanations following Kamalaśīla’s pivotal victory in debate over a Chinese rival of the Ch’an tradition (discussed below). The ‘doctrine’ of Bhk 1 is described as that of the three kinds of wisdom (śrutamayī, cintāmayī, and bhāvanāmayī prajñā). The way of meditation of Bhk 2 is explained in light of the realization that there is only one vehicle; it is the result of this meditation that Bhk 3 is said to explain. But such categorical statements are best made with caution; all three texts contain discussions of doctrine, meditation, and its result. 2. Perhaps it is as much due to the excellence of scholarship already devoted to their study as it is to the breadth of their concern that the Bhāvanākramas tend to be among the most widely quoted of Indian Buddhist texts. Tucci has provided critical editions of the Sanskrit and Tibetan of the Bhk 1 (1958) and the Sanskrit of the Bhk 3 (1971). Of the three texts, the original Sanskrit of the Bhk 2 is lost. As well, the fi rst folio of the Sanskrit of Bhk 1 is missing, as are the edges of many of the pages of the manuscript of Bhk 3 from which Tucci worked. All three texts are, however, fully preserved in the Tibetan Tanjur. A critical edition of the Tibetan text of Bhk 2 based on the Narthang (N), Peking (P), Derge (D), and Cone (C) editions has been prepared by K. Goshima (1983). The Central Institute of Higher Tibetan Studies has published an edition of the Sanskrit and Tibetan texts, which occasionally serves to clarify Tucci (Namdol, 1997). This contains a Hindi translation and Sanskrit reconstructions of Bhk 2 and the fi rst folio of Bhk 1. I have worked mainly from the editions of Tucci, Goshima, and the Derge Tanjur dBu ma KI (22a- 41b, 42a-55b, and 56a-68b respectively for the three texts). Unless otherwise noted, references are to Tucci for the Sanskrit texts and to D for the Tibetan.

The three texts contain numerous instructions for the beginner in Mahāyāna meditation. Equally, the Bhāvanākramas constitute a kind of apology or justifi cation for a particular approach to the Buddhist path. The Tibetan tradition regards them as containing a summary of arguments employed in the refutation of a Chinese Ch’an position being advocated at the time of the fi rst great transmission of Buddhism to Tibet. The debate has been characterized in terms of gradualism vs. subitism (Gomez, 1987). The gradualist view, associated with the Indian side led by Kamalaśīla, held that Awakening can only be attained after a long process of training in which one deliberately cultivates certain causes and conditions conducive to its occurrence. These causes and conditions are both moral and cognitive – one must cultivate specifi c moral virtues as well as a specifi c conceptual knowledge of the nature of reality. Such cultivation (bhāvanā) is a gradual process – it takes time and has defi nite steps. The subitist position, represented by a Ch’an monk (Hvashang) named Mo ho yen (Sanskrit: Mahāyāna), held that Awakening occurs suddenly, all at once. Awakening was understood as a state requiring only the practice of a non-conceptual concentration or absorption (dhyāna), wherein one’s mind is cleared from all obscuring mental activity. Attempts to cultivate specifi c moral virtues and views of reality were understood as counterproductive on the grounds that they accumulate karma and prolong one’s sojourn through cyclical existence.3 The contrary view, argued by Kamalaśīla, held that a particular kind of cognitive process – a ‘correct analysis’ or ‘discernment of reality’ (bhūta-pratyavekṣā) – is essential to the achievement of Awakening. Because Awakening involves a kind of knowledge (i.e. non-conceptual knowledge, nirvikalpa jñāna), and not merely concentration, it is essential to fi rst become established not only in concentration but also in a correct conceptual knowledge, which can then function to give rise to the sought after noetic state. The principle at work here is that like arises from like: one kind of knowledge arises on the basis of another. Kamalaśīla seems to have understood his opponent as arguing on the basis of the same causal principle, but focusing on the other aspect of Awakening – its non-conceptuality. Thus,

3. Bhk 3 13.15–14.1: yas tu manyate / cittavikalpa samutthāpitaśubhāśubha-karmavaśena sattvāḥ svargādi karmaphalam anubhavantaḥ saṃsāre saṃsaranti / ye punar na kiṃcic cintayanti nāpi kiṃcit karma kurvanti te parimucyante saṃsārāt / tasmān na kiṃcic cintayitavyam / nāpi dānādikuśalacaryā kartavyā / kevalaṃ murkhajanam adhikṛtya dānādikuśalacaryā nirdeṣteti /; D 61b1: gang zhig sems kyi rnam par rtog pas bskyed pa’i dge ba dang mi dge ba’i las kyi dbang gis sems can rnams mtho ri la sogs pa’i ‘bras bu myong zhing ‘khor ba na ‘khor ro / gang dag ci yang mi sems ci yang mi byed pa de dag ni ‘khor ba las yongs su thar bar ‘gyur ro / de lta bas na ci yang mi bsam mo / sbyin pa la sogs pa dge ba spyad par yang mi bya’o / sbyin pa la sogs pa spyod pa ni skye bo blun po’i dbang du mdzad nas bstan pa kho na yin no snyom du sems shing de skad kyang smra ba des ni theg pa chen po mtha dag spangs pa yin no/: ‘But some consider, “Because they are subject to positive and negative actions generated by the conceptual mind, sentient beings spin around in cyclical existence experiencing the fruits of their actions, such as heaven. But those who do not think anything nor perform any action whatsoever, they are fully liberated from cyclical existence. Therefore nothing should be thought. Nor should the skillful conduct of giving and the rest be undertaken. The skillful conduct of giving and the rest is taught only with foolish people in mind”’.

© Equinox Publishing Ltd 2006 74 BUDDHIST STUDIES REVIEW as a non-conceptual state of knowledge, Awakening might be thought of as only arising on the basis of non-conceptual concentration. According to Kamalaśīla, this is a misunderstanding; non-conceptual concentration, because it lacks a cognitive dimension, can not on its own result in a state of knowledge. At the same time, however, Kamalaśīla did recognize the concentrative nature of the resulting state of nonconceptual knowledge; he therefore accepted the necessity of initially combining the one-pointed quality of concentration with the noetic quality of conceptual knowledge. The resulting state could thus be both concentrated and noetic.4 To understand Kamalaśīla’s views in more detail, I will attempt to demonstrate how he understood the logical relations obtaining between bhāvanā and dhyāna, as well as their relationships to other key terms denoting meditative states and processes. I will then attempt to demonstrate how it is that Kamalaśīla accepted as normative the concept of bhāvanā. But before entering into these topics it would perhaps be germane to say a few words about how I understand the English word ‘meditation’. In normal English usage, and in its most general conception, when one talks of ‘meditation’, in most cases one is referring to a deliberately undertaken introspective process which is aimed at reaching a qualitatively diﬀ erent state of mind – usually a spiritual state of some description (e.g. communion with God) or a heightened state of awareness. The process itself is marked by concentration – either upon some aspect of the goal sought or upon the activity itself. Such concentration usually follows a technique, which can be described and practised. Although introspective, this may involve a physical aspect. Practices of meditation vary widely, including everything from visualization, repetition of verbal phrases or prayers, to the walking of labyrinths. These diverse procedures share the features of voluntariness, introspection and concentration, and are all undertaken with the aim of bringing about an altered state of consciousness or a change in spiritual condition. It is important to note, however, that in the western intellectual tradition there exists a second and related use of the word ‘meditation’ in which many of these features are not found. In this case the word meditation is employed to refer to processes of ordinary rational thought that are seriously undertaken and concerned with topics judged to be important or profound. ‘Meditation’ in this sense is a kind of intellectual contemplation or rumination, involving neither a special technique of concentration nor the idea of achieving of an altered state of consciousness. This employment of the word is perhaps most famously exempliﬁ ed

4. Thus the two opponents both asserted that an initial practice of concentration was necessary, but they disagreed as to its nature. Just as from Kamalaśīla’s perspective, Mo ho yen’s diﬃ culty was to explain the noetic aspect of Awakening on the basis of a non-cognitive practice; from Mo ho yen’s perspective, Kamalaśīla’s diﬃ culty would be to explain Awakening’s nonconceptuality arising on the basis of a conceptual process. In addition, as noted, Mo ho yen held that such conceptual activities were karmatic and thus counterproductive with respect to liberation.

in the title of René Descartes’ Meditationes de Prima Philosophia (1641). Clearly, in western culture there is a long-standing and distinct association between ‘meditation’ and ordinary discursive thought. With these considerations in mind, we may now return our attention to the idea of meditation as found in the Bhāvanākramaḥ texts. It may be remarked in passing that, in their own way, these three works of Kamalaśīla have shown themselves to be as seminal to the Tibetan Buddhist tradition as the reﬂ ections of the great French philosopher have proven to western philosophy. And the questions they raise appear equally profound. The issue we are concerned with here, namely the relations obtaining among diverse concepts of meditation, is an extraordinarily complex one. An initial listing of some of the main terms for meditation occurring within the Bhāvanākramas includes the following (I provide some of the more common English translations):

1. dhyāna – absorption, trance, concentration, meditation 2. samādhi – concentration, state of concentration, meditation 3. bhāvanā – cultivation, development, realization, actualization, meditation 4. śamatha – tranquillity meditation, serenity, calm, calm abiding 5. vipaśyanā – insight meditation, insight, wisdom

It can be seen that there are, in fact, at least three Sanskrit terms commonly translated by the English word ‘meditation’, plus two that are considered kinds of meditation. In order to understand Kamalaśīla’s particular views, I will fi rst of all briefl y review and comment upon each of these. Dhyāna – probably the term most commonly associated with the word ‘meditation’, this word is derived from the Sanskrit verbal root √dhyai (to think of, meditate upon) plus the primary affi x (kṛt-pratyaya) ‘lyuṭ’, which forms abstract, instrumental or verbal nouns. In Chinese, as is well known, the transliterated term Ch’an (Japanese: Zen) came to refer to a number of schools emphasizing the meditation aspect of Buddhist practice. In the original Indian context, however, the word specifi cally referred to a set of four successive meditative absorptions (Pāli: jhāna) wherein one’s mental contents are attenuated to a state of one-pointedness and equanimity free from all disturbing emotions and mental activity.5 Although much has been written about the dhyānas, for our purposes three points should be noted. First, it is important to recognize that after the fi rst dhyāna, all thought is said to have been eliminated. Both vitarka (relatively gross thought, as mental application) and vicāra (subtle thought, or examination) are absent from the second through to the fourth dhyānas. While some thought is present in the fi rst dhyāna, this state is not defi ned as one in which a deliberate conceptual analysis of reality takes place.

5. The dhyānas are said to be optionally followed by another sequence of four ‘formless attainments’ (ārūpya samāpatti) occurring on the basis of the fourth dhyāna. See Griﬃ ths (1986), Cran- gle (1994: 201–7), and Gethin (1998: 184–6) for treatments of these attainments.

The second point is that Kamalaśīla subdivides the fi rst absorption into two.6 The fi rst division contains both vitarka and vicāra, the second contains vicāra but not vitarka. This second division he calls ‘intermediate absorption’ (dhyānāntara).7 We shall see that Kamalaśīla may actually have accepted the possibility of a deliberate conceptual analysis of reality occurring in the fi rst dhyāna; if so, it might well have been the case that he particularly associated this possibility with the intermediate dhyānā. We shall return to this topic at the end of this paper. In any case, for now, it seems clear that he regarded his opponent as adhering to a general notion of dhyāna that excludes deliberate activities of conceptual analysis. The third point to note is that the dhyānas are all said to share the quality of ekagratā or one-pointedness of mind. This quality is also said to characterize our next term, samādhi, as well as śamatha. Samādhi - this term is often given as a gloss for dhyāna, but its scope is generally understood to be wider than that of the four dhyānas. While it encompasses these, it also refers to other states of mental one-pointedness. For example, it includes the preliminary state of ‘the capable’ (anāgamya – equivalent to ‘access’

6. Bhk 1 209.2–11: eṣā ca cittaikāgratā uttarottarakarmaṇyatāsaṃprayogād ālambanādiguṇaviśeṣayogā c ca dhyānārūpya[samāpatti]vimokṣādivyapadeṣaṃ labhate / tathā hi yadopekṣāvedanāsaṃprayuktā savitarkasavicārā sā bhavati / tadānāgamyam ucyate / yadā ca kāmatṛṣṇayā [pāpadharmaiḥ] viviktā bhavati [vitarkavicāra]prītisukhādhyātmasaṃprasādaiḥ saṃprayuktā bhavati / tadā prathamam dhyānam ucyate / ata eva prathamaṃ dhyānaṃ vitarkamātrarahitaṃ dhyānāntaram ucyate / yadā vitarkavicārarahitā prathamadhyānabhūmitṛṣṇayā viviktā ca bhavati / prītisukhādhyātmasaṃprasāda iḥ saṃprayuktā bhavati / tadā dvitīyaṃ dhyānam ucyate /; D 32b 2–5: sems rtse gcig pa de phyi phyir las su rung ba nyid dang ldan zhing dmigs pa la sogs pa’i yon tan gyi khyad par thob pas bsam gtan dang / gzugs med pa’i snyoms par ‘jug pa dang / rnam par thar pa la sogs pa’i ming thob po / ‘di ltar gang gi tshe btang snyoms kyi tshor ba dang ldan zhing rtog pa dang bcas pa dang / dpyod pa dang bcas par gyur pa de’i tshe mi lcogs pa med pa zhes bya ste / [ bsam gtan dang po’i sbyor ba’i sems so ] / gang gi tshe ‘dod pa’i sred pa dang / sdig pa’i chos rnams dang bral zhing rtog pa dang / dpyod pa dang / dga’ ba dang / bde ba dang ldan par gyur pa de’i tshe bsam gtan dang po zhes bya’o / bsam gtan dang po de nyid las rtog pa tsam med pa de ni bsam gtan khyad par can zhes bya’o / gang gi tshe rtog pa dang dpyod pa dang bral te bsam gtan dang po’i sa’i sred pa dang bral bar gyur nas dga’ ba dang bde ba dang / nang yongs su dang ba dag dang ldan par gyur pa de’i tshe bsam gtan gnyis pa zhes bya’o/: ‘And this one-pointedness of mind receives the designation, “absorption”, “formless attainment”, “liberation”, and so forth on account its being endowed with greater and greater capability and on account of possess- ing the distinct qualities of its object and so forth. That is, when it is conjoined with feelings of equanimity, and has gross thought and subtle thought, then it is called “capable” (anāgamya, T. mi lcogs pa med pa, “not unable”). And when it is separated from the thirst for pleasure [as well as] conjoined with [gross and subtle thought,] joy, happiness, and internal clarity (ādhyātma- saṃprasād, i.e. mindfulness and clear comprehension), then it is called the fi rst absorption. After this the fi rst absorption without gross thought alone is called “intermediate absorption” (dhyānāntara). When it is without (both) gross thought and subtle thought, as well as separated from thirst for the stage of the fi rst absorption – (but still) conjoined with joy, happiness, and clarity pertaining to the self – then it is called the second absorption’. 7. In including an intermediate stage between the fi rst and second dhyānas, one in which vitarka is absent while vicāra remains, the account of the dhyānas found in the Bhāvanākramas appears to be following that of Vasubandu. See Abhk viii 22d, 23d-e. Like Kamalaśīla, Vasubandhu divides the fi rst dhyāna into two. However, according to Vasubandhu, the fi rst dhyāna may contain either vitarka or vicāra, but not both in the same moment.

(upacāra) concentration in Theravāda Buddhism), a degree of mental focus that allows one to enter into the dhyānas proper, or else to turn one’s attention to insight (vipaśyanā) practices. The word samādhi is thus arguably the broadest Indian term used in reference to states of meditation. In the context of spiritual practice, it is understood to exclude non-virtuous states of concentration, such as those an assassin and so forth. It usually denotes concentration as a mental state of nondistraction. As well, the word is sometimes used to refer to the processes that give rise to such mental states.8 Derived from the verbal root √dhā (to put, place, set, join, unite) in combination with the prefi xes ‘sam’ (together) and ‘ā’ (around) and the masculine suffi x ‘ki’, the sense of samādhi is one of ‘placing (that which is) around together’, or simply, ‘unifi cation’. Bhāvanā - derived from the causative form of the verbal root √bhū, (√bhū + ṇic + either lyu or lyuṭ), the word bhāvanā literally means ‘causing to be’, ‘making become’, or ‘giving rise to’. Hence among its most common translations we fi nd ‘cultivation’, ‘development’, ‘realization’, and even ‘actualization’. These translations are etymologically more precise than ‘meditation’; in addition, the fi rst two have the extra implication of a gradual progression toward a state that is sought. Unlike the term samādhi, one does not properly speak of ‘attaining’ a state of bhāvanā; on the contrary, this word exclusively refers to processes through which one brings particular states into being. It should be noted, however, that translations such as ‘cultivation’ and ‘development’, while capturing this primary signifi cation, do not necessarily imply samādhi. Outside of the context of spiritual practice, the word bhāvanā obviously can have a wider, non-technical sense that includes deliberately giving rise to things not necessarily characterized by concentration, nor indeed consistent with virtue. For example, the activities of watering a garden, cooking a meal, and plotting an act of revenge can all be considered as instances of bhāvanā in a broader sense of ‘cultivation’. In the context of spiritual practice, however, this word refers to virtuous eff orts, those that further the cause of liberation by generating positive states (dharmas) characterized by concentration (samādhi).9

8. See e.g. Visuddhimagga III 2–3, where the idea of a process is captured in its gloss: samadhāna. (Note again the kṛt-pratyaya ‘lyuṭ’ in a word indicating a process). The two uses of the word can be seen clearly in connection with the classical eightfold path, where samādhi not only appears as one of the eight aspects of the path, namely, right concentration (samyak samādhi), but also as the label for the set of three aspects that together constitute the meditation component of the path: right eff ort, right mindfulness, and right concentration. The term is thus used to designate the meditation (samādhi) component of the three reliances of morality, meditation, and wisdom (śīla, samādhi, and prajnā), into which the eightfold path may be subdivided. On such use of the term samādhi to designate a list of items (though not necessarily items related to meditation) see Skilton (2002). 9. According to Vasubandhu, bhāvanā is divisible into four types corresponding to a classical fourfold division of right eff ort – eff orts for the arising of non-arisen pure dharmas, for the growth of already arisen pure dharmas, for the non-arising of non-arisen impure dharmas and for the destruction of already arisen impure dharmas. Abhk 1081.1-5, 22–31. For a detailed treat- ment of the classical Nikāya account see Gethin (1992: 69–80). In terms of this term’s relation to

Śamatha: in the Buddhist tradition bhāvanā is generally understood to be divisible into the two subcategories of tranquillity (śamatha) and insight (vipaśyanā). Kamalaśīla accepts this division.10 The term śamatha (Pāli samatha) is derived from the verbal root √śam (to be quiet, to cease, to rest).11 The principal signifi cations of śamatha are those of calmness and the capacity to remain continuously focused on one object of meditation. Thus the cultivation of tranquillity brings about states of concentration and calm, such as the dhyānas. With respect to Awakening, the function of śamatha is to stabilize the mind, thereby making vipaśyanā possible. Vipaśyanā: this term is the Buddhist hybrid Sanskrit word corresponding to the Pāli vipassanā. It is derived from the verbal root √pas ‘to see’, plus the prefi x vi which can have the senses of ‘apart, asunder’ and ‘diff erent, distinct’. The resulting sense is one of ‘seeing into’ or ‘discerning’. Hence ‘insight’ is the usual translation for this term. In general, vipaśyanā is understood to refer to observational and analytic processes that lead to a knowledge of reality. In the Bhāvanākramas, vipaśyanā is specifi cally identifi ed with a technical term, namely, ‘the discernment of reality’ (bhūta-pratyavekṣā). As Kamalaśīla quotes from the Sūtras, ‘Tranquillity is one-pointedness of mind; insight is the discernment of reality’.12 The function of vipaśyanā is to perceive the elements of reality (dharmas) as they truly are. If the eff ect of śamatha is to enable vipāśyanā, it is vipaśyanā that allows for non-conceptual knowledge to occur. And on this basis Awakening is gradually achieved. By the power of tranquillity the mind becomes steady on its object, like a lamp [burning] in a place without wind. By insight, the light of correct

samādhi, one might note that the process of making such eff orts would involve concentrating (samādhāna) on the desired state. In addition, when specifi c states of concentration (such as those of dhyāna) are aimed at, this might be thought of as a case of the fi rst right eff ort, that which is aimed at the arising of non-arisen pure dharmas. This, however, was apparently a point of controversy between diff erent Buddhist schools. The Vaibāśikas apparently considered samādhi a separate mental dharma while the Sautrāntikas thought it simply referred to a concentrated mind (Abhk 1126.6-1127.3). 10. Bhk 2 D46b1–2: rnal ‘byor pas ni sgom pa’i dus thams cad du nya dang sha la sogs pa spang zhing mi mthun pa ma yin pa dang / zas tshod zin par bza’ bar bya’o // de ltar byang chub sems dpa’ zhi gnas dang lhag mthong gi tshogs mtha’ dag bsags pa des bsgom pa la ‘jug par bya’o //: ‘The yogin, forsaking meat and fi sh at all times of meditation (sgom pa, bhāvanā), should eat only the proper amount of food and that which is not incompatible (with the scriptures). In this manner, bodhisattvas who have accumulated all the conditions of tranquillity and insight (zhi gnas dang lhag mthong gi tshogs, śamatha-vipaśyanā-saṃbhāra) should enter into meditation’. 11. Plus the kṛṭ-pratyaya ‘athac’. See Pāṇini 3.3.92. Thanks to Sanjay Kumar Shastri of McGill Univer- sity for clarifying the derivations of samādhi and śamatha. 12. Bhk 3 3.1–4: tatra śamathaś cittaikāgratā / vipaśyanā bhūtapratyavekṣeti saṃkṣepād āryaratnameghādau bhagavatā śamathavipaśyanayor lakṣanam uktam /; D 56b3–4: de la mdor na zhi gnas ni sems rtse gcig pa nyid do / lhag mthong ni yang dag pa la rtog pa’o / zhes bcom ldan ‘das kyis ‘phags pa dkon mchog sprin la sogs pa las zhi gnas dang lhag mthong gi mthsan nyid bka’ stsal to /: ‘Thus in the noble Ratnamegha and elsewhere the Bhagavān concisely stated the defi nition of tranquillity and insight, “Tranquillity is one-pointedness of mind, insight is the discernment of reality”’. Also quoted at Bhk 2 D 47a2.

knowledge emerges on the basis of accurately realizing the true nature of dharmas. And on that basis all obscuration is removed, just as the night by the dawning of the sun.13 The diff erence between tranquillity and insight can also be understood in terms of the application of concepts to the object of the meditation. Adhering to the Saṃdhinirmocana Sūtra, Kamalaśīla asserts that śamatha is nonconceptual (nirvikalpa) and that vipaśyanā is conceptual (savikalpa). [T]he Bhagavān taught four realities as meditation objects for yogis: a) a refl ection without conceptualization b) a refl ection accompanied by conceptualization c) the limit of things and d) the perfection of purpose. In this context, when by means of tranquillity one has committed oneself to a refl ection of all dharmas or to a form like that of the Buddha, that which is depended upon is called a refl ection without conceptualization (nirvikalpa-pratibimbakam). It is called without conceptualization here because of an absence of concepts determining the real object-meaning (bhūtārtha). And it is called a refl ection because it is depended upon, having committed oneself to a refl ection of dharmas as they have been learned and understood. When, by means of insight, the yogin analyses (vicārayati) that very refl ection in order to realize reality, then it is called a refl ection accompanied by conceptualization (savikalpa-pratibimbakam) on account of the presence there of a concept determining reality (tattva), which is the characteristic of insight.14

13. Bhk 3 1.10–14: śamathabalena svālambane cittam aprakampyaṃ bhavati nivātasthitapradīpavat / vipaśyanayā yathāvad dharmatattvāvagamāt samyagjñānālokaḥ samutpadyate / tataḥ sakalam āvaraṇaṃ prahīyate / andhakāravad ālokodayāt / D 56a2–3: zhi gnas kyi stobs kyis ni mar me rlung med pa na gnas pa bzhin du dmigs pa las sems mi gyo bar ‘gyur ro / lhag mthong gis mun par nyi ma shar ba bzhin du chos kyi de kho na ji lta ba bzhin du rab tu rtogs pas yang dag pa’i shes rab kyi snang ba ‘byung ste / sgrib pa mtha dag spong bar ‘gyur ro /. 14. Bhk 3 1.14–2.5: / ata eva bhagavatā catvāry ālambanavastūni yogināṃ nirdiṣṭāni / nirvikalpapratibimbakam / savikalpapratibimbakam / vastuparyantatā / kāryapariniśpattiś ca / tatra śamathena yat sarvadharmapratibimbakaṃ buddhādirūpaṃ cādhimucyālambyate tan nirvikalpapratibimbakam ucyate / tatra bhūtārthanirūpaṇāvikalpābhāvān nirvikalpakam ucyate / yathāśrutodgṛhītānāñ ca dharmāṇāṃ pratibimbakamadhimucyālambyata iti kṛtvā pratibimbakam ucyate / tad eva pratibimbakaṃ yadā vipaśyanayā vicārayati yogī tattvādhigamārthaṃ tadā savikalpapratibimbakam ucyate / tattvanirūpaṇāvikalpasya vipaśyanālakṣaṇasya tatra samudbhavāt /; D 56a3–7: de lta bas na bcom ldan ‘das kyis rnal ‘byor pa rnams kyi dmigs pa’i dngos po bzhi bstan te / rnam par mi rtog pa’i gzugs brnyan dang / rnam par rtog pa dang bcas pa’i gzugs brnyan dang / dngos po’i mtha dang / dgos pa yongs su grub pa’o / de la zhi gnas kyis na chos thams cad kyi gzugs brnyan gang yin pa dang / sangs rgyas kyi gzugs la sogs pa la mos nas dmigs pa ste / de ni rnam par mi rtog pa’i gzugs brnyan zhes bya’o / de la yang dag pa’i don la rnam par rtog pa med pas na de rnam par mi rtog pa zhes bya’o / ji ltar thos pa dang ji ltar zin pa’i chos rnams kyi gzugs brnyan zhes bya’o / rnal ‘byor pas de kho na’i don rtogs par bya ba’i phyir gang gi tshe lhag mthong gis gzugs brnyan de nyid la spyod pa de’i tshe lhag mthong gi mtshan nyid de kho na la rtogs pa’i rnam par rtog pa de na yod pas na rnam par rtog pa dang bcas pa’i gzugs brnyan zhes bya’o /.

Thus, according to Kamalaśīla, in vipaśyanā concepts (vikalpa) are deliberately applied when one analyses (vicārayati) the meditation object. Kamalaśīla’s use of the verbal form, vicārayati, can be taken to indicate the fact that he considered a kind of subtle thought (vicāra) to be present in vipaśyanā. If this is so, it raises the question as to Kamalaśīla’s views regarding the compatibility of vipaśyanā and dhyāna. Before addressing this question directly, it would be prudent to investigate Kamalaśīla’s understanding of the logical relations obtaining among the other key meditation terms. An important passage that captures these relations can be found at the opening of the third Bhāvanākramaḥ: Homage to Tārā! The Bhāvanākramaḥ is related in brief for those who are beginners in the way of the Mahāyāna Sūtras. In that context, even if the samādhi of bodhisattvas was taught by the Bhagavan to be limitless, by way of the (four) immeasurables and all the rest, nevertheless all samādhis are subsumed under tranquillity and insight. Therefore, precisely that path which carries the union of tranquillity and insight is related.15 There are many points made in this brief passage. First, the universe of discourse is defi ned: as we have noted, the text is specifi cally addressing the spiritual practice of beginners who wish to become profi cient in the way set out in the Mahāyāna Sūtras. Secondly, for our purposes, we must notice that every samādhi or state of concentration is here said to be included under the rubric of tranquillity and insight. It is notable that the division of tranquillity and insight is made among samādhis or states of concentration, and not within bhāvanā as such. Given the fact that bhāvanā is also divisible into tranquillity and insight, this consideration suggests a conception of samādhi as coextensive with bhāvanā within the intended universe of discourse for these texts. Assuming for the moment that this is so, we can imagine Kamalaśīla’s working understanding of the relationship between the terms as follows. Like the concepts of lake and lakeshore, or parent and child, the concepts of samādhi and bhāvanā would be mutually impli- cative, although not identical in meaning. The diff erence in their usages would rest in the fact that bhāvanā is a term for meditative processes while samādhi is a term principally used to indicate meditative states, as well as processes.16 If this is so, then Kamalaśīla’s working understanding of the relationship between

15. Bhk 3 1.2–6: namas tārāyai / mahāyānasūtrāntanayapravṛttānāṃ saṃkṣepato bhāvanākramaḥ kathyate / tatra yady api bodhisattvānām aparimito ‘pramāṇādibhedena bhagavatā samādhir upadiṣṭaḥ, tathāpi śamathavipaśyanābhyāṃ sarve samādhayo vyāptā iti / sa eva śamathavipaśyanā-yuganaddhavāhī mārgas tāvat kathyate / D 56b6–57a2: ‘jam dpal gzhon nur gyur pa la phyag ‘tshal lo / theg pa chen po mdo sde’i tshul la zhugs pa rnams kyi phyir bsgom pa’i rim pa mdor brjod par bya’o / de la bcom ldan ‘das kyi byang chub sems dpa’ rnams kyi ting nge ‘dzin tha dad pa tshad med dpag tu med pa la sogs pa bstan du zin kyang / zhi gnas dang lhag mthong gnyis kyis ting nge ‘dzin thams cad la khyab pas na zhi gnas dang lhag mthong zun du ‘brel pa ‘jug pa’i lam de nyid brjod par bya’o /. 16. That bhāvanā is the principal term for meditative processes is reﬂ ected in the fact that it is typi- cally a conjugation of the verbal root √bhū (e.g. bhāvayet, one should meditate on, one should

samādhi and bhāvanā is as follows: any instance of bhāvanā implies samādhi and any instance of samādhi implies bhāvanā (Bh if and only if S). In eﬀ ect, what this means is that we cannot conceive one without also implying the other. (It does not mean that every proposition that is true of the one is true of the other). Given this understanding, we may now return to the question of Kamalaśīla’s understanding of the relationship between dhyāna and vipaśyanā. We can approach this topic initially by asking how our author may have regarded the relationship between dhyāna and bhāvanā. While Kamalaśīla nowhere comments directly on this issue, given the analysis just made, samādhi can be seen as a kind of ‘bridge term’ linking his conceptions of dhyāna and bhāvanā. By recalling the relationship between samādhi and dhyāna, discussed earlier, a logical structure begins to emerge. The four dhyānas, it will be remembered, are all speciﬁ c forms of meditative concentration (samādhi). Thus any instance of dhyāna is also an instance of samādhi (If Dh then S). The reverse, however, is not the case; we have seen that samādhi is the wider term, encompassing some forms of concentration not included in the four dhyānas. From this it follows that while dhyāna implies bhāvanā (If Dh then Bh), the reverse (If Bh then Dh) is not the case. There are some instances of bhāvanā where dhyāna is not involved. With these considerations in mind, we can view our question concerning the interrelationship of meditation terms in the Bhāvanākramas as a kind of dilemma of translation. If, on the one hand, we translate dhyāna as meditation, then this would open the door to the logical possibility that some forms of bhāvanā would not be properly conceived of as meditation for these texts. This would be consistent with a view of bhāvanā as ‘cultivation’ in the wider, non-technical sense mentioned above. Some instances of bhāvanā would not have been considered by Kamalaśīla as involving that concentration or one-pointedness of mind which, in the Buddhist tradition, is the hallmark of meditative states of consciousness. In particular, this way of understanding the texts opens up the possibility that the vipaśyanā component of bhāvanā might have been conceived as a kind of complimentary intellectual process of logical reasoning (yukti) that is not fundamen- tally meditative in nature. Vipaśyanā might not have been understood as a form of meditative concentration in the technical sense of one pointedness of mind. According to this way of thinking, the term ‘meditation’ would be restricted to states of samādhi (including the dhyānas) in which there is no deliberate discursive activity.17

cultivate) that is employed when the actual procedure for meditating is being described. There are no instances of a conjugation of sam + ā + √dhā in these texts. In the Tibetan, as well, it is the verb sgom pa (equivalent to bhāvanā) that is employed in these contexts. 17. This way of reading the Bhāvanākramas would emphasize the continuity of Kamalaśīla’s thought with that of the Buddhist epistemological tradition going back to Dignāga. Hayes (1988: 168) and Prévèreau (1994: 33) have both suggested that cintāmayī prajñā is identiﬁ ed as a kind of vipaśyanā meditation by Dignāga (c. 480–540). According to Prévèreau, ‘Yaśomitra suggests that vipaśyanā is synonymous with prajñā (AKIV:14) so that there is not only the insight brought

This understanding, however, raises a problem. If vipaśyanā, as a process of conceptual analysis, is not understood as meditative in nature, this would appear to be at odds with Kamalaśīla’s acceptance of the subdivision of samādhi into non- conceptual śamatha and conceptual vipaśyanā. We would be forced to conclude that Kamalaśīla’s account is inconsistent. If, on the other hand, bhāvanā is translated as ‘meditation’, then the process of vipaśyanā will necessarily be understood as meditative in nature. However, the nature of vipaśyanā as a mental process that is at once concentrative and analytic will be problematic. How can the mind remain focused on one point and engage in conceptual analysis at the same time? On this understanding, Kamalaśīla’s account would appear to be unintelligible. So how do we decide? In order to address this issue, I will discuss Kamalaśīla’s ideas in light of a fundamental conceptual paradigm upon which the Bhāvanākramas are based. I will then argue that if we wish to take this deep structure of Indian thought seriously, the best term to translate as ‘meditation’ is bhāvanā. On this account, Kamalaśīla’s conception of vipaśyanā would be one of a meditative process, rather than one of ordinary logical reasoning.

THREE KINDS OF WISDOM

The arguments found in the Bhāvanākramas rest upon a deep cultural presuppo- sition that there exist three basic kinds of wisdom or learning (prajñā) – those associated with study, thinking, and meditation (śrutamayī prajnā, cintāmayī prajñā, and bhāvanāmayī prajñā). This paradigm constitutes one of India’s oldest and most fundamental ways of conceiving the process of spiritual practice.18 It is already mentioned in the Pāli Canon and later in Buddhaghosa’s Visuddhimagga and Vasubandhu’s Abhidharmakośa.19 Although conceptions of its scope and place on

about by mental discipline (bhāvanā) to which today’s meditators tend to limit themselves, but also that brought about by critical reasoning (cintā)’. Prévèreau asserts that Dignāga intro- duced a new way of doing vipaśyanā consisting in critical reasoning. It could well be that a ten- dency to identify insight with profound (but nevertheless non-meditative) thinking originated with Dignāga. Even so, the infl uences upon Kamalaśīla were many and he may well have held to a conception of meditation based more upon the descriptions given in various sūtras, particularly those of the Yogācāra school. This is the light in which I have interpreted his views. 18. For recent treatments, see Balagangadhara (2005: 1005–7) and especially, for Vasubandhu and Kamalaśīla, Nichols (2005). 19. See Dīgha Nikāya III 219; Vism XIV, 14: 438. Interestingly, what appears to be a parallel version of the same schema is to be found in the Bṛhadāraṇyaka Upaniṣad, where the terminology dif- fers but the general threefold structure is clearly identifi able B.U. II 4.5: ātmā va are draṣṭavyaḥ śrotavyo mantavyo nididhyāsitavyaḥ / maitreyi ātmano vāre darśanena śravaṇena matyā vijñānenedaṃ sarvaṃ viditam //: ‘You see, Maitreyī – it is one’s self (ātman) which one should see and hear, and on which one should refl ect and concentrate. For by seeing and hearing one’s self, and by refl ecting and concentrating on one’s self, one gains the knowledge of this whole world’ (trans. in Olivelle, 1996: 28–9). See also B.U. IV 5.6 (69–70). I am grateful to Christine Fillion for giving me these references. See Fillion (2004).

© Equinox Publishing Ltd 2006 ADAM KAMALAŚĪLA’S BHĀVANĀKRAMAS: A PROBLEM OF TRANSLATION 83 the Buddhist path vary, among western students of Buddhism this prototype has today become so generally accepted as to border on the platitudinous rather than the profound. Its signifi cance, however, may not have always been fully appreci- ated. When examined, this apparently simple schema yields some rather curious and intriguing implications. Here we shall begin our examination with one of its early formulations – the account found in Vasubandhu’s Abhidharmakośa: (The wisdom) arisen from study (śrutamayī) is a certitude born from the authority of a qualifi ed person (āpta-vacana-prāmāṇyajāta); the wisdom arisen from thinking (cintāmayī) is born from the profound refl ection of reasoning (yukti-nidhyānajā); the wisdom arisen from bhāvanā (bhāvanāmayī) … is born from concentration (samādhijā).20 The account given by Vasubandhu regards the three kinds of wisdom as pro- gressive – the wisdom arisen from thinking basing itself on the wisdom arisen from study, and the wisdom arisen from bhāvanā basing itself on that arisen from thinking.21 Kamalaśīla describes the wisdom of thinking as a process in which one disentangles two kinds of scriptural statement, the nītārtha and the neyārtha. This important hermeneutical distinction is a tool developed by Buddhists in order to classify scriptural statements as respectively defi nitive or provisional in

20. Abhk 892.3–4: āptavacanaprāmāṇyajātaniścayā śrutamayī, yuktinidhyānajā cintāmayī, samādhijā bhāvanāmayīti … . 21. Abhk 861.4–6: śrutvā cintayati / aviparītam cintayitvā bhāvanāyāṃ prayujyate / samādhau tasya śrutamayīṃ prajñāṃ niśritya cintāmayī jāyate / cintāmayīṃ niśritya bhāvanāmayī jāyate /: ‘Having studied, he thinks. Having thought correctly, he engages in bhāvanā. Having relied on the wisdom arisen from study of one in samādhi, (the wisdom) arisen from thinking is born. Having relied on (the wisdom) arisen from thinking, (the wisdom) arisen from bhāvanā is born’. It is important to notice that Vasubandhu interprets the affi x (taddhita-pratyaya) -mayaṭ in a very specifi c way, namely, as indicating a cause (hetu) and thus having a sense of ‘arisen from’ or ‘transformation of’. The idea of ‘cause’ appears to be intended here in the sense of a preceding, eff ective cause. Abhk 892.4–5: tadyathā – annamayāḥ prāṇāḥ, tṛṇamayyo gāva iti / : ‘Vital breaths are the result of food, cows are the result of grass’. However, it is also normal to understand the suffi x more in the sense of a present, material cause meaning ‘made of’, ‘consisting of’, or even ‘having the nature (svabhāva) of’. See Pāṇini 4.3.82 ff . The ambiguity here may be understood to imply another important ambiguity: that between ‘state’ and ‘process’. If we accept the understanding of -mayaṭ as indicating a preceding eff ective cause, then the wisdoms associated with study, thinking and bhāvanā are more readily conceived ‘statically’ as results, which is to say as states of knowledge that have arisen. On the other hand, if we take the affi x to indicate a cause in the sense of a present material cause (‘consisting in’, ‘having the nature of’) then the three kinds of wisdom may also be interpreted in a more ‘dynamic’ sense, as processes. The affi x -mayaṭ is ambiguous and lends itself to being interpreted in either one of the two senses noted, depending on context. Another way of making the same point would be to say that it can function analogously to either a pañcāmī tatpuruṣa (abla- tive determinative) or a karmadhārya compound. On the latter interpretation, wisdom would be identifi ed with the very processes of studying, thinking and bhāvanā, not with the results of these processes. It appears to me that Kamalaśīla’s understanding shifts between these two conceptions. I therefore translate -mayaṭ in an open ended manner, analogous in meaning to a ṣāṣṭhī tatpuruṣa – i.e. the wisdom of thinking.

© Equinox Publishing Ltd 2006 84 BUDDHIST STUDIES REVIEW meaning. For the Madhyamaka tradition, the distinction is cashed out in terms of their reference: defi nitive statements are those that pertain to ultimate truth (i.e. variously anutpāda, śūnyatā, parama-tattva, tathatā, pudgala-dharma-nairātmya, etc.), provisional statements refer to the conventional (Thurman, 1978: 26, 32–4). According to Kamalaśīla’s account, it is the task of the wisdom of thinking to identify which statements refer to the real object or meaning (bhūtam arthaṃ) and which do not. The task of the wisdom of bhāvanā is to realize the meaning or object that is real. There, fi rst of all, the wisdom of study should be generated. For through it one initially enters into the meaning of the scriptures. Thereafter one penetrates their provisional and defi nitive meanings by the wisdom of thinking. After that, having ascertained the meaning that is real (bhūta, i.e. nītārtha) by means of that (wisdom of thinking), one should realize (bhāvayet) it, not that which is unreal (abhūta i.e. neyārtha).22 Thus Kamalaśīla’s discussion of the wisdom of bhāvanā details the process for experientially realizing the conclusions regarding ultimate reality that have already been reached through the wisdom of thinking. It is intriguing to observe that the same verbal root √bhū lies at the basis of both bhāvanā and the word we have here translated as ‘real’, bhūta. We should also notice another important conceptual link made by Kamalaśīla himself in this context: that between bhāvanā and another term derived from √bhū, namely, anubhāva, ‘experience’ (literally, ‘following the real’). In the fi rst Bhāvanākramaḥ, Kamalaśīla commences his discussion of bhāvanāmayī prajñā with the statement, ‘Experience (anubhāva) belongs to those who practice’.23 This kind of experiential wisdom can be seen to depend on and encompass the other two kinds of wisdom. One ‘realizes’ or ‘meditates on’ the conclusions already reached by thinking. What it means to realize or meditate in this way is a rather subtle question. While it defi nitely includes a conceptual dimension, it also appears to be diff erent from a simple case of thinking deeply about some profound topic. The wisdom of bhāvanā is conceived as having a ‘direct’ character, it is ‘experiential’ – this is what distinguishes it from the mere wisdom of thinking (cintāmayī

22. Bhk 1 198.10–13: tatra prathamaṃ tāvat śrutamayī prajñotpādanīyā / tayā hi tāvad āgamārtham avadhārayati / tataś cintāmayyā prajñayā nītaneyārthaṃ nirvedhayati / tatas tayā niścitya bhūtam arthaṃ bhāvayen nābhūtam /. Also see Bhk 2 D 46a7–b1: tshul bzhin bsam pa gang zhe na / gang nges pa’i don gyi mdo sde dang drang ba’i don gyi mdo sde la sogs pa legs par gtan la ‘bebs pa ste / de ltar byang chub sems dpa’ the tshom med na bsgom pa la gcig tu nges par ‘gyur ro // de lta ma yin na the tshom gyis ‘phyang mo nyug pa’i theg pa la ‘dug pa ni lam kha brag gi mdor phyin pa’i mi ltar gang du yang gcig tu nges par mi ‘gyur ro //: ‘And who is correct in thought? He is one who is settled with regard to the system of the Sūtras of deﬁ nitive and provisional meaning. If a bodhisattva is without doubt in this regard, he will be certain in meditation. If it was not like this, if one was on an uncertain course due to doubt, then he couldn’t become certain of anything at all – just like a man who has arrived at the juncture of a forked road’. 23. Bhk 1 204.14–15: anubhāvaś ca pratipattṝṇāṃ. D 30b 3: sgrub pa po rnams la yang nyams yod de /.

prajñā). It is an experiential process of discerning reality, one that occurs in a concentrated state (samādhi). In the Buddhist context, direct experience possesses an epistemologically privileged position; it is considered indubitable in a way that conclusions reached through reasoning alone are not. Here the two forms of wisdom may be con- trasted as ‘thinking through’ what one has studied (cintāmayī prajñā) versus actually ‘going through’ or concentratedly ‘experiencing’ the reality of what one has already thought through (bhāvanāmayī prajñā). In the third Bhāvanākramaḥ, Kamalaśīla makes this point using an analogy: And whatever is known through the wisdom of study and thinking is itself to be realized through the wisdom of bhāvanā (bhāvanāmayyā prajñayā bhāvanīyaṃ), nothing else. (For example), it is like a horse running along a previously indicated running track. Therefore the discernment of reality (bhūta-pratyavekṣā) is to be undertaken.24 Thus Kamalaśīla maintains that bhāvanāmayī prajñā is required for Awakening and that this form of wisdom constitutes a more direct realization than cintāmayī prajñā. From this he concludes that the ‘discernment of reality’ (bhūta-pratyavekṣā) should be undertaken. It will be recalled that this ‘discernment of reality’ is explicitly identifi ed by Kamalaśīla with insight (vipāśyanā). Thus it seems that bhāvanāmayī prajñā, conceived as a process for realizing the true nature of reality, can here be especially identifi ed with one branch of the well-known two-branch schema of bhāvanā: śamatha and vipaśyanā. As the most profound of the three kinds of wisdom, bhāvanāmayī prajñā is identifi able with insight par excellence. Given this fact it would seem reasonable to translate bhāvanā as ‘meditation’ in the context of these texts. The problem with translating dhyāna as ‘meditation’ is that Kamalaśīla’s opponent is portrayed as adhering to a conception of dhyāna that does not include processes of bhūta- pratyavekṣā. While Kamalaśīla himself may have accepted the idea that vipaśyanā can occur in the fi rst dhyāna, his opponent is portrayed as defi nitely not accepting any such conjunction. How then are we to understand the relation between the two concepts of meditation for these texts? The answer is that bhāvanā is best conceived as the broader term: bhāvanā includes dhyāna –a term which is principally associated with non-conceptual meditation – but it is not exhausted by it. If this is so, then dhyāna would perhaps be translated as ‘absorption’, or as some other term sugges- tive of its status as a subspecies of meditation. After all, one would not normally wish to translate ‘pomme’ as ‘fruit’!

24. Bhk 3 20.3–6: kiṃ ca yad eva śrūtacintāmayyāprajñayā viditaṃ tad eva bhāvanāmayyā prajñayā bhāvanīyaṃ nānyat / saṃdiṣṭa[dhāvana]-bhūmyaśvadhāvanavat / tasmāt bhūtapratyavekṣā kartavyā / D 64a2–3: yang thos pa dang bsam pa las byung ba’i shes rab kyis rtogs pa gang yin pa de nyid bsgom pa las byung ba’i shes rab kyi bsgom par bya’i / gzhan du ni ma yin te rta dkyu sa kyi sa bstan nas rgyug pa bzhin no / de lta bas na yang dag par so sor brtag par bya’o /. On the analogy of the running horse, see Abhk 328: 10–13.

We can see that the question of the best Sanskrit equivalent for ‘meditation’ in the Bhāvanākramas is not unrelated to the debate between Kamalaśīla and Mo ho yen. It is perhaps not without reason that these three texts were so repetitively entitled ‘The Process of Bhāvanā’. Bhāvanā is a term for processes that include the development of wisdom through concentrated conceptual activity. In the Bhāvanākramas, Kamalaśīla portrays his opponent as adhering to a conception of dhyāna that excludes deliberate conceptual activity. Kamalaśīla’s charge against his Ch’an rival consisted precisely in the claim that he failed to understand the necessity of conceptual activity in the achievement of Awakening. As an advocate of dhyāna, Mo ho yen was viewed as interpreting Awakening as an accomplishment achieved simply by ceasing all mental activity.25 But according to Kamalaśīla, it is only through the particular conceptual activity that is the discernment of reality (bhūta-pratyavekṣā, vipaśyanā, prajñā-bhāvanā) that nonconceptual knowledge or gnosis (nirvikalpa-jñāna) can arise.26 While Kamalaśīla’s criticisms of Mo ho yen are centred upon the idea of knowledge (jñāna), Mo ho yen’s critique of Kamalaśīla can be viewed as focusing on the idea of action (karma). It is the deliberate, volitional nature of the conceptual activities enjoined by Kamalaśīla that he objects to. Volitional activity is precisely that which binds sentient beings to the wheel of rebirth. As such, it is counterproductive. This would seem to be the crux of the disagreement. For Kamalaśīla, some actions are necessary to the achievement of Awakening. If, then, bhāvanā is to be considered the broader term for meditation in these texts, and if vipaśyanā is a kind of bhāvanā that is necessary for Awakening, we

25. This doctrine is ascribed to the Ājīvakas. Bhk 3 20.14–16: yac cāpy ucyate / na kiṃcit kuśalādikarma kartavyam iti / tatraivaivaṃvadatā karmakṣayān muktir ity ājivakavādābhyupagamo bhavet / ; D 64b1 4: yang dag ba la sogs pa’i las ci yang mi bya’o zhes zer ba de ni de skad smra bas las zas nas grol bar ‘gyur ro zhes mu stegs can kun tu tshol ba’i smra ba khas blang par ‘gyur ro /: ‘Now as for what is also said – that not a single action, skillful or otherwise, should be performed – those who speak thus would here be accepting the doctrine of the Ājīvakas, that is, liberation on the basis of karma’s destruction’. 26. When the practitioner reaches the point of comprehending emptiness nonconceptually, this constitutes ‘the limit of things’ mentioned above (Bhk 3 2.8–10; D 56a7–b1) and the arising of the fi rst stage and transcendent path of the bodhisattva. On this basis, gradually but inevitably the bodhisattva’s purpose is perfected and the omniscience of Buddhahood is achieved. Quot- ing from the Ratnakuṭa, the ultimate justifi cation for the practice of insight is dramatically explained. Bhk 2 D 49b5–b6: gang shes rab kyis dngos po’i ngo bo nyid so sor brtags nas mi bsgom gyi / yid la byed pa yongs su spong ba tsam ‘ba’ zhig sgom par byed pa de’i rnam par rtog pa nam yang mi ldog (NP rtog) cing ngo bo nyid med pa nyid (NP omit nyid) rtogs (Goshima follows NP: rtog) par yang mi ‘gyur te / shes rab kyi snang ba med pa’i phyir ro // ‘di ltar “yang dag par so sor rtog pa nyid las yang dag pa ji lta ba bzhin du shes pa’i me byung na gtsubs shing gtsubs pa’i me bzhin du rtog pa’i shing sreg go” zhes bcom ldan ‘das kyis bka’ stsal to //: ‘Someone who only cultivates the mere abandonment of mental activity, but who does not meditate having analysed the nature of entities with wisdom, will never get rid of concepts and will not come to realize the absence of inherent nature – on account of the absence of the light of wisdom. So it is said by the Illustrious One, “When the fi re of knowing reality as such arises from the very discernment of reality, it incinerates the wood of concepts, just as the fi re of fi resticks rubbed together [consumes the sticks themselves]”‘. See also Bhk 3 30.8–11.

© Equinox Publishing Ltd 2006 ADAM KAMALAŚĪLA’S BHĀVANĀKRAMAS: A PROBLEM OF TRANSLATION 87 may well ask what precisely its undertaking was thought to involve. Here I can only give a brief indication of Kamalaśīla’s conception, in relation to other meditation terminology already discussed. First of all, the process is described as being undertaken while actually abiding in a state of śamatha. ‘[H]aving renounced all obscurations, one who wants pure knowledge to arise must cultivate wisdom while abiding in tranquillity’.27 Similar considerations apply to samādhi. Quoting from the Saṃdhinirmocana Sūtra: … [H]aving abandoned mental distractions, he inwardly discerns those very same previously considered dharmas as refl ections in the sphere of concentration (T. ting nge ‘dzin, Skt. samādhi). In this manner, discriminating the meaning of what is to be known in those refl ections in the sphere of concentration, thoroughly discriminating, completely consid- ering, completely investigating, forbearing, accepting, classifying, looking and knowing – That is called insight. So it is that the bodhisattva is skilled in insight.28 Thus while Kamalaśīla’s views regarding the compatibility of vipaśyanā with both samādhi and śamatha are clear, the question still remains as to whether he regarded its conjunction with dhyāna as possible. In the Bhāvanākramas the two terms are never mentioned in the same breath. In spite of this fact, my suggestion is that Kamalaśīla did regard them as compatible and that, given the presence of thought (vitarka-vicāra) within the fi rst dhyāna, it is precisely this meditative state that theoretically allows the two to come together. Indeed, among the dhyānas, this conjunction would have been considered possible only in the fi rst dhyāna – since thought is absent from the second to the fourth dhyānas. In particular, it may well have been the higher, intermediate division of the fi rst absorption (dhyānāntara) that Kamalaśīla associated with the possibility of the practice of insight meditation. It will be recalled that it is in this division that gross thought (vitarka) is absent while subtle thought (vicāra) remains. If we associate the activity of subtle thought with the verbal form vicārayati, employed by Kamalaśīla in

27. Bhk 2 D 44b7–45a1: de lta bas na sgrib pa ‘mtha dag spangs nas yongs su dag pa’i ye shes ‘byung bar ‘dod pas zhi gnas la gnas shing shes rab bsgom par bya’o //. While basic, such a notion has been taken by some scholars as suggesting a conceptual tension in Buddhist meditation theory. How can conceptual analysis occur in a state of one-pointed meditation? According to Griﬃ ths, it led to various attempts to regard insight as occurring in ‘liminal states’ between the dhyānas. This diﬃ culty may well provide some explanation for the postulation of an intermediate dhyāna. (1983: 245–51, 285–7; also see Vetter, 1988: xxv–xxvii). 28. Bhk 2 D 47a7–47b2: sems kyi rnam par g.yeng ba spangs nas ji ltar bsams pa’i chos de dag nyid nang du ting nge ‘dzin gyi spyod yul gzugs brnyan du so sor rtog par byed / mos par byed do // de ltar ting nge ‘dzin gyi spyod yul gzugs brnyan de dag la shes bya’i don de rnam par ‘byed pa dang / rab tu rnam par ‘byed pa dang / yongs su rtog pa dang / yongs su dpyod pa dang / bzod pa dang / ‘dod pa dang / bye brag ‘byed pa dang / lta ba dang / rtog pa gang yin pa de ni lhag mthong zhes bya ste / de ltar na byang chub sems dpa’ lhag mthong la mkhas pa yin no” zhes gsungs so //. (See Powers, 1995: 150–52, 341–2).

© Equinox Publishing Ltd 2006 88 BUDDHIST STUDIES REVIEW describing the practice of insight according to the Saṃdhinirmocana Sūtra, the connection would be made.29 The other possibility, as discussed, is that insight be understood in terms of

29. Indeed this would seem to be confi rmed by Kamalaśīla’s employment of the same verb in describing the experiential process of conceptual analysis outlined in the Laṅkāvatāra Sūtra. In editing the Sanskrit text of Bhk 1, Tucci created a separate section for this description, No. 16, which he entitled ‘Method of meditation according to the Laṅkāvatāra; vicāra on the dharmas (no object, no subject), etc’. This section occurs immediately following Kamalaśīla’s discussion of dhyāna; both are set in the overall context of bhāvanāmayī prajñā. Although it is beyond the scope of this paper to deal extensively with Kamalaśīla’s account of the Laṅkāvatāra Sūtra, a brief excerpt should suffi ce to demonstrate the experiential quality of the language he employs. After having stabilized the mind on the fi ve aggregates as a meditation object (Bhk 1 206.7–15), the analytic process is described as commencing with an analysis of dharmas with material form: Bhk 1 210.16–211.4: prathamaṃ yogī ye rūpiṇo dharmā bāhyārthatayā paraiḥ parikalpitās teṣu tāvad vicārayet / kim ete vijñānād anye, āhosvid vijñānam evaitat tathā pratibhāsate, yathā svapnāvasthāyām iti / tatra vijñānād bahiḥ paramāṇuśo vicārayet / paramāṇūmś ca bhāgaśaḥ pratyavekṣamāṇo yogī tān arthān na samanupaśyati / tasyāsamanupaśyata evaṃ bhavati / cittamātram evaitat sarvaṃ na punar bāhyo ‘rtho vidyate / tad evam / “cittamātraṃ samāruhya bāhyam arthaṃ na kalpayet ” rūpidharmavikalpān tyajed ityarthaḥ / teṣām upa[labdhi]lakṣaṇaprāptānāṃ vicārayed anupalabdheḥ / evaṃ rupiṇo dharmān vibhāvyārūpiṇo vibhāvayet /; D 33a4–34b1: thog mar rnal ‘byor pas chos gzugs can gang dag gzugs la sogs pa phyi rol gyi don du gzhan dag gis brtags pa de dag la ci ‘di dag rnam par shes pa las gzhan zhig yin nam / ‘on te rnam par shes pa de nyid de ltar snang ste / rmi lam gnas skabs ji lta ba bzhin nam zhes dpyad par bya’o / de la rnam par shes pa las phyi rol pa rdul phra rab tu bshig ste / rdul phra rab rnams kyang cha shas kyis so sor brtags na rnal ‘byor pas don de dag mi mthong ngo / des de dag ma mthong bas ‘di snyam du ‘di dag thams cad ni sems tsam ste phyi rol gyi don med do snyam du sems so / ‘di ltar / sems tsam la ni rab brten nas / phyi rol don la mi brtag go / zhes de skad ‘byung ba ni chos gzugs can la rnam par rtog pa spong ba’o zhes bya ba’i tha tshig go / dmigs su rung ba’i mtshan nyid du gyur pa de dag rnam par dpyad na mi dmigs pa’i phyir ro / de ltar chos gzugs can rnams rnam par bshig nas gzugs can ma yin pa rnam par bshig par bya ste /: ‘First of all the yogin should analyse (vicārayet, T. dpyad par bya) those dharmas having a material form, imagined by others as being external objects: “Are these other than consciousness, or is it this consciousness itself appearing in that manner – just as in dreamstate?” In that regard [i.e. if the position held is that they have a nature] outside of consciousness, he should break them down into atoms (paramāṇuśo vicārayet, T. rdul phra rab tu bshig ste). And discerning (pratyavekṣamāṇaḥ, T. so sor brtags pa) those atoms by way of parts, the yogin does not see (na samanupaśyati, T. mi mthong) those things. Not seeing (them), he thinks: “All this is indeed mind-only, an external object does not exist”. Therefore thus: “Having ascended to mind-only, one would not imagine an external object”. The meaning is that he would abandon conceptualizations of dharmas that have a material form. He should draw a conclusion (vicārayet, T. rnam par dpyad) from the non- apprehension of those things that are in principle apprehensible. Thus having broken down (vibhāvya, T. rnam par bshig nas) dharmas with a material form, he should break down (vibhāvya, T. rnam par bshig bya) those without material form’. It is apparent that here the conceptual analysis or ‘breaking down’ of experienced realities is considered part of the process of insight. Nevertheless, it would be a mistake to consider the inferences comprising this analysis as instances of cintāmayī prajñā. Because they are undertaken while in the sphere of samādhi, they are properly considered as meditative in nature; they form part of what is meant by bhāvanāmayī prajñā. They appear to be distinct from cases of ordinary inference insofar as they seem to be conceived as directly ‘based upon’ objects and objective states of aff airs being concurrently experienced in meditation. This appears to be so in the sense that the meditator is thought of as being able to remain one-pointedly focused upon such objects, holding them in view while ‘analysing’ them.

non-experiential processes of ordinary reasoning (cintāmayī prajñā); this opens the door to two possible ways of translating. The fi rst would take dhyāna as the default term for meditation, exclusively referring to states in which there is no deliberate discursive activity. Vipaśyanā would be understood as a complimentary intellectual process that is not meditative in nature. But the problem with this suggestion is that it does not recognize Kamalaśīla’s acceptance of vipaśyanā as a subdivision of samādhi, as discussed above. A second, more sophisticated possibility would treat both nonconceptual dhyāna and rational vipaśyanā as kinds of meditation – albeit forms which are distinct and mutually exclusive in their natures. As rational insight, vipaśyanā would count as a kind of meditation much in the same way as do Descartes’ refl ections for the western intellectual tradition. On this account, the process of meditation would have to consist of a serial alternation, back and forth, between the modes of ordinary rational thought and wholly non-conceptual concentration.30 While coherent, the problem with this account is that it fails to take seriously the Indian division of wisdom into three kinds and the clear connection between vipaśyanā and bhāvanāmayī prajñā. Furthermore, and perhaps more tellingly, it does not accurately refl ect Kamalaśīla’s own descriptions of the process of insight. A careful reading of the texts shows that Kamalaśīla’s understanding of the discernment of reality is not that of a kind of ordinary reasoning, but that of a special kind of conceptual analysis that occurs while actually abiding in a state of meditative concentration (samādhi). While it is true that Kamalaśīla nowhere explicitly states that this includes the fi rst dhyāna, there is no reason in principle to exclude this possibility, given the presence of thought therein. In fact, given Kamalaśīla’s descriptions of the analytic process, the upper ‘intermediate’ division of the fi rst dhyāna seems the most likely candidate for the sphere of concentration in which this special kind of conceptual analysis is practised. As described in the Bhāvanākramas, the discernment of reality involves cultivating an accurate perception of the true nature of the con- stituents of conventional reality. In other words, it involves the ‘discrimination of dharmas’ (dharma-pravicaya). This discrimination involves mindfulness practices (smṛtyupasthāna) and specifi c acts of what might be called ‘perceptual judgement’ as to the ultimate emptiness of dharmas. At one point Kamalaśīla describes this experiential process of examining the nature of dharmas as being so vivid that it is like looking at the blemishes on one’s face through its refl ection in a mirror.31 It

30. On the conception of the process of meditation as a kind of serial alternation, see Williams (1989: 72–4) and Ruegg (1989: 111–12). 31. Bhk 3 2.5–8: tasyaiva ca pratibimbasya svabhāvaṃ nirūpayan yogī, darpaṇāntargatasvamukhap ratibimbapratyavekṣaṇena svamukhagatavairūpyāṇāṃ viniścayavat, sarvadharmāṇāṃ yathāvat svabhāvāgamāt /; D 56a7–8: rnal ‘byor pa gzugs brnyan de nyid kyi ngo bo nyid la rtog pas chos thams cad kyi ngo bo nyid ji lta ba bzhin du khong du chud de / bdag gis bzhin gyi gzugs me long gi nang du byung ba la brtags na bdag gi bzhin la mi sdug pa la sogs pa mngon pa bzhin no /: ‘And in determining the nature of that very reﬂ ection on the basis of understanding the nature of all dharmas as they are, the yogin is as if ascertaining blemishes upon his own face by discerning its reﬂ ection in a mirror’.

© Equinox Publishing Ltd 2006 90 BUDDHIST STUDIES REVIEW seems clear that Kamalaśīla is not describing a case of ordinary logical reasoning, but rather a subtle form of meditative analysis. It is an intentionally undertaken practice that occurs in a heightened state of one-pointed consciousness, a practice that is at once conceptual analysis and meditation.

The author would like to thank David Higgins of the University of Victoria for reviewing an earlier version of this paper, with a special emphasis on the Tibetan passages. As well, a special acknowledge- ment is owed to Prof. Peter Harvey for his insightful comments and helpful suggestions in preparing this paper for publication.

ABBREVIATIONS

Abhk Abhidharmakośa & bhāṣya of Ācarya Vasubandhu with Sphutārthā commentary of Ācarya Yaśomitra, Swami Dwārikādās Śāstrī (ed.) (Varanasi: Bauddha Bharati, 1987). B. U. Bṛhadāraṇyaka Upaniṣad Bhk Bhāvanākramaḥ Pāṇini Aṣṭadhyāyī of Pāṇini, Roman transliteration and English trans. Sumitra M. Katre (1987) (Delhi: Motilal Banarsidass, 1989). Skt. Sanskrit Vism Visuddhimagga T. Tibetan Editions of the Tibetan Tanjur: C – Cone; D – Derge; N – Narthang; P - Peking

BIBLIOGRAPHY

Primary sources not cited from western style editions

Bsgom pa’i rim pa (bhāvanākramaḥ). sDe dGe Tibetan Tripitaka bsTan ‘gyur. No. 3915 [22a1-41b7], No. 3916 [42a1-55b5], No. 3917 [55b6-68b7]. Faculty of Letters, University of Tokyo, 1979.

Printed books and articles

Adam, Martin T., 2002. ‘Buddhism and Hermeneutics’. Focus Pragensis, vol. II, pp. 95–113. Balagangadhara, S. N., 2005. ‘How to Speak for the Indian Traditions: An Agenda for the Future’. Journal of the American Academy of Religion, vol. 73(4), pp. 987–1013. Conze, Edward, 1973. Buddhist Thought in India. Ann Arbor, MI: University of Michigan Press. Conze, Edward, 1975. Further Buddhist Studies: Selected Essays. Oxford: Bruno Cassirer. Crangle, Edward Fitzpatrick, 1994. The Origin and Development of Early Indian Contemplative Practice. Wiebadan: Harrassowitz Verlag. Fillion, C. M., 2004. The Role of Scriptural Testimony, Reason and Spiritual Practice in The Upadeśasāhasrī: A Non-commentarial Work of Śaṅkara. Unpublished Masters thesis, McGill University. Gethin, Rupert, 1992. The Buddhist Path To Awakening. Leiden: E.J. Brill. Gethin, Rupert, 1998. The Foundations of Buddhism. Oxford: Oxford University Press. Gomez, Louis O., 1983. ‘Indian Materials on the Doctrine of Sudden Enlightenment’. In Early Ch’an in China and Tibet, Whalen Lai & Lewis R. Lancaster (eds), pp. 393–434. Berkeley, CA: Asian Humanities Press.

Gomez, Louis O., 1987. ‘Purifying Gold: The Metaphor of Eff ort and Intuition in Buddhist Thought and Practice’. In Sudden and Gradual: Approaches to Enlightenment in Chinese Thought, Peter N. Gregory (ed.), pp. 67–165. Delhi: Motilal Banarsidass. Goshima, Kiyotaka, 1983. The Tibetan Text of the Second Bhāvanākramaḥ. Kyoto (self-published, Showado printing). Griffi ths, Paul J., 1983. ‘Indian Buddhist Meditation Theory: History, Development and Systematization’. Unpublished PhD dissertation, University of Wisconsin-Madison. Griffi ths, Paul J., 1986. On Being Mindless. Delhi: Sri Satguru Publications. [Reprint 1999.] Hayes, Richard, 1988. Dignāga on the Interpretation of Signs. Dordrecht: Kluwer. Kajiyama, Yuichi, 1991. ‘Later Mādhyamikas on Epistemology and Meditation’. In Mahāyāna Buddhist Meditation: Theory and Practice, Minoru Kiyota (ed.), pp. 114–43. Delhi: Motilal Banarsidass. Klein, Anne, 1986. Knowledge and Liberation: Tibetan Buddhist Epistemology in Support of Transformative Religious Experience. Ithaca, NY: Snow Lion Publications. Lamotte, Étienne, 1935. Le Samdhinirmocanasutra: L’explication des Mystères. Louvain: Université de Louvain. Lopez, Jr., Donald S., 1987. A Study of Svātantrika. Ithaca, NY: Snow Lion Publications. Lopez, Jr., Donald S., 1993. ‘On the Interpretation of the Mahāyāna Sūtras’. In Buddhist Hermeneutics, D. Lopez Jr. (ed.). Delhi: Motilal Banarsidass. Ñāṇamoli, Bhikkhu, 1991. The Path of Purifi cation (Visuddhimagga of Buddhaghosa). Kandy: Buddhist Publication Society. [Reprint.] Namdol, Gyaltsen, 1997. Bhāvanākramaḥ of Ācārya Kamalaśīla (Tibetan Version, Sanskrit Restoration and Hindi Translation). Sarnath: Central Institute of Higher Tibetan Studies. Nichols, Brian, 2005. ‘Elaborations of the Threefold Typology of Wisdom from Vasubandhu to Kamalaśīla: A Consideration of the Relationship between Refl ection and Meditation’. Unpublished paper delivered at the Eastern Division Meeting of the American Philosophical Association (International Society of Buddhist Philosophy subpanel), New York. Olivelle, Patrick, 1996. Upaniṣads. Oxford: Oxford University Press. Pagel, Ulrich, 1995. The Bodhisattvapiṭaka: Its Doctrines and their Position in Mahāyāna Literature. Tring: The Institute of Buddhist Studies. Powers, John, 1995. Wisdom of Buddha: The Saṃdhinirmocana Sūtra. Berkeley, CA: Dharma Publishing. Prévèreau, Raynald, 1994. ‘Dharmakīrti’s Account of Yogic Intuition as a Source of Knowledge’. Unpub- lished Masters thesis, McGill University. Ruegg, David Seyfort, 1989. Buddha-Nature, Mind, and the Problem of Gradualism in a Comparative Perspective. London: School of Oriental and African Studies. Ruegg, David Seyfort, 1992. ‘On the Tibetan Historiography and Doxography of the “Great Debate at bSam yas”’. In Tibetan Studies: Proceedings of the 5th Seminar of the International Association for Tibetan Studies 1989, pp. 237–44. Narita: Naritasan Shinshoji. Ryusei, Keira, 1999. Mādhyamika and Epistemology: A Study of Kamalaśīla’s Method for Proving the Voidness of All Dharmas; Introduction, Annotated Translations and Tibetan Texts of Selected Sections of the Second Chapter of the Madhyamakāloka. Unpublished PhD thesis, Université de Laussane. Schmithausen, Lambert, 1976. ‘On the Problem of the Relationship of Spiritual Practice and Philo- sophical Theory in Buddhism’. In German Scholars on India Vol. II, pp. 235–50. Bombay: Nachiketa Publications. Schmithausen, Lambert, 1981. ‘On Some Aspects of Descriptions or Theories of “Liberating Insight” and “Enlightenment” in Early Buddhism’. In Studien Zum Jainismus und Buddhismus, pp. 199–250. Hamburg: Franz Steiner Verlag GmbH Wiesbaden. Skilton, Andrew, 2002. ‘State or Statement? Samādhi in Some Early Mahāyāna Sutras’. The Eastern Buddhist, vol. 34(2), pp. 51–93. Sopa, Geshe, 1991. ‘Śamathavipaśyanāyuganaddha: The Two Leading Principles of Buddhist Meditation’. In Mahāyāna Buddhist Meditation: Theory and Practice, Minoru Kiyota (ed.), Delhi: Motilal Banarsidass, pp. 46–65. Taniguchi, Fujio, 1992, ‘Quotations from the First Bhāvanākramaḥ of Kamalaśīla Found in Some Indian Texts’. In Tibetan Studies: Proceedings of the 5th Seminar of the International Association for Tibetan

Studies 1989, pp. 303–7. Narita: Naritasan Shinshoji. Thurman, Robert, 1978. ‘Buddhist Hermeneutics’. Journal of the American Academy of Religion, vol. 46, pp. 19–39. Tucci, Giuseppe, 1958. Minor Buddhist Texts, Part II, First Bhāvanākramaḥ. Rome: Serie Orientale Roma ix. 2/ Delhi: Motilal Banarsidass. [Reprinted 1986.] Tucci, Giuseppe, 1971. Minor Buddhist Texts, Part III, Third Bhāvanākramaḥ. Roma: Instituto Italiano per il Medio Ed Estremo Oriente. Vaidya, P. L. (ed.), 1961. Samādhirājasūtra. Darbhanga: The Mithila Institute. Vetter, Tilmann, 1988. The Ideas and Meditative Practices of Early Buddhism. Leiden: E.J. Brill. Wallace, B. Alan, 1998. The Bridge Of Quiescence: Experiencing Tibetan Buddhist Meditation. Chicago, IL: Open Court. Williams, Paul, 1989. Mahāyāna Buddhism. London: Routledge. Williams, Paul, 1991. ‘On the Interpretation of Madhyamaka Thought’. Journal of Indian Philosophy, vol. 19, pp. 191–218.